
import pandas as pd
import numpy as np
df=pd.read_excel('Student Applications & Performance.xlsx')


df.head()
df.shape

(3400, 56)


#FUNCTION USED INORDER TO SEPERATE NUMERICAL AND CATEGORICAL DATA


%run dqr.py


report=dataQuality(df)


type(report)

dict


report.keys()

dict_keys(['numeric', 'categorical'])


report['numeric']


##SORTING OUT THE EARLY ATTRITION STUDENTS


att=df[df['RETURNED_2ND_YR']==0]

att


import seaborn as sb


df['RETURNED_2ND_YR'].value_counts()

1    2677
0     723
Name: RETURNED_2ND_YR, dtype: int64


### 723 NON RETURNEES (21%) FROM A BATCH OF 3399 STUDENTS


bygen_famcontribution=df.groupby('RETURNED_2ND_YR')['EST_FAM_CONTRIBUTION'].mean().reset_index()
bygen_famcontribution


by_unmetneeds=df.groupby('RETURNED_2ND_YR')['UNMET_NEED'].mean().reset_index()


by_unmetneeds


p=sb.barplot(x='RETURNED_2ND_YR',y='UNMET_NEED',data=by_unmetneeds)
for i,v in by_unmetneeds.iterrows():
    p.text(v.name,v.UNMET_NEED,round(v.UNMET_NEED,2),ha='center')


by_highsclgpa=df.groupby('RETURNED_2ND_YR')['HIGH_SCHL_GPA'].mean().reset_index()
by_highsclgpa


by_attempthrs=df.groupby('RETURNED_2ND_YR')['FIRST_TERM_EARNED_HRS'].mean().reset_index()
by_attempthrs


by_DISTANCE_FROM_HOME=df.groupby('RETURNED_2ND_YR')['DISTANCE_FROM_HOME'].mean().reset_index()
by_DISTANCE_FROM_HOME


h=sb.barplot(x='RETURNED_2ND_YR',y='DISTANCE_FROM_HOME',data=by_DISTANCE_FROM_HOME)
for i,v in by_DISTANCE_FROM_HOME.iterrows():
    p.text(v.name,v.DISTANCE_FROM_HOME,round(v.DISTANCE_FROM_HOME,2),ha='center')


by_STDNT_TEST_ENTRANCE_COMB=df.groupby('RETURNED_2ND_YR')['STDNT_TEST_ENTRANCE_COMB'].mean().reset_index()
by_STDNT_TEST_ENTRANCE_COMB


report['categorical']


by_STDNT_GENDER=att.groupby('STDNT_GENDER')['RETURNED_2ND_YR']


by_STDNT_GENDER.value_counts()

STDNT_GENDER  RETURNED_2ND_YR
F             0                  475
M             0                  248
Name: RETURNED_2ND_YR, dtype: int64


by_IN_STATE_FLAG=att.groupby('IN_STATE_FLAG')['RETURNED_2ND_YR']
by_IN_STATE_FLAG.value_counts()

IN_STATE_FLAG  RETURNED_2ND_YR
N              0                   89
Y              0                  634
Name: RETURNED_2ND_YR, dtype: int64


by_HOUSING_STS=att.groupby('HOUSING_STS')['RETURNED_2ND_YR']
by_HOUSING_STS.value_counts()

HOUSING_STS  RETURNED_2ND_YR
Off Campus   0                  389
On Campus    0                  334
Name: RETURNED_2ND_YR, dtype: int64


by_STDNT_MAJOR=att.groupby('STDNT_MAJOR')['RETURNED_2ND_YR']
by_STDNT_MAJOR.value_counts()

STDNT_MAJOR                     RETURNED_2ND_YR
Accounting                      0                   12
Applied Computer Science        0                    5
Art                             0                   18
Art Education                   0                    2
Biology                         0                   82
Biology and Secondary Ed        0                    3
Chemistry                       0                   10
Communication                   0                   14
Computer Science - Games        0                   14
Computer Science - Systems      0                    9
Criminal Justice                0                   29
Early Childhood Education       0                   37
Earth and Space Science         0                    1
Engineering Studies             0                   15
English Language/Literature     0                    9
English and Secondary Ed        0                    4
Exercise Science                0                   19
Finance                         0                    8
French                          0                    1
French with Teacher Cert        0                    1
General Business                0                   12
General Studies/AS              0                    2
Health Science                  0                    6
Health and Physical Education   0                    5
History                         0                    8
History and Secondary Ed        0                    7
Information Technology          0                    3
Management                      0                   15
Management Information Systems  0                    7
Marketing                       0                   15
Mathematics                     0                    6
Mathematics and Secondary Ed    0                    1
Middle Grades Education         0                    3
Music                           0                    6
Music Education                 0                    8
Music Performance               0                   12
Nursing                         0                   37
Political Science               0                   14
Pre-Business                    0                   32
Pre-Engineering/RETP            0                    7
Pre-Nursing                     0                   47
Psychology                      0                   43
Sociology                       0                    4
Spanish with Teacher Cert       0                    1
Spec Ed: Gen. Curr. - Reading   0                    3
Theatre Arts                    0                   12
Theatre Education               0                    3
Undeclared                      0                  111
Name: RETURNED_2ND_YR, dtype: int64


by_FATHER_HI_EDU_DESC=att.groupby('FATHER_HI_EDU_DESC')['RETURNED_2ND_YR']
by_FATHER_HI_EDU_DESC.value_counts()

FATHER_HI_EDU_DESC         RETURNED_2ND_YR
College/Beyond             0                  257
High School                0                  289
Middle School/Junior High  0                   26
No Information             0                  103
Other/Unknown              0                   48
Name: RETURNED_2ND_YR, dtype: int64


by_MOTHER_HI_EDU_DESC=att.groupby('MOTHER_HI_EDU_DESC')['RETURNED_2ND_YR']
by_MOTHER_HI_EDU_DESC.value_counts().plot(kind='barh',color='blue')

<AxesSubplot:ylabel='MOTHER_HI_EDU_DESC,RETURNED_2ND_YR'>


by_MOTHER_HI_EDU_DESC.value_counts()

MOTHER_HI_EDU_DESC         RETURNED_2ND_YR
College/Beyond             0                  301
High School                0                  265
Middle School/Junior High  0                   16
No Information             0                  123
Other/Unknown              0                   18
Name: RETURNED_2ND_YR, dtype: int64


by_STDNT_BACKGROUND=att.groupby('STDNT_BACKGROUND')['RETURNED_2ND_YR']
by_STDNT_BACKGROUND.value_counts().plot(kind='barh',color='purple')

<AxesSubplot:ylabel='STDNT_BACKGROUND,RETURNED_2ND_YR'>


by_STDNT_BACKGROUND.value_counts()

STDNT_BACKGROUND  RETURNED_2ND_YR
BGD 1             0                  477
BGD 2             0                   24
BGD 3             0                  170
BGD 4             0                   26
BGD 5             0                    4
BGD 6             0                   14
BGD 7             0                    7
BGD 8             0                    1
Name: RETURNED_2ND_YR, dtype: int64


df.columns

Index(['STUDENT IDENTIFIER', 'STDNT_AGE', 'STDNT_GENDER', 'STDNT_BACKGROUND',
       'IN_STATE_FLAG', 'INTERNATIONAL_STS', 'STDNT_MAJOR', 'STDNT_MINOR',
       'STDNT_TEST_ENTRANCE1', 'STDNT_TEST_ENTRANCE2',
       'STDNT_TEST_ENTRANCE_COMB', 'FIRST_TERM', 'CORE_COURSE_NAME_1_F',
       'CORE_COURSE_GRADE_1_F', 'CORE_COURSE_NAME_2_F',
       'CORE_COURSE_GRADE_2_F', 'CORE_COURSE_NAME_3_F',
       'CORE_COURSE_GRADE_3_F', 'CORE_COURSE_NAME_4_F',
       'CORE_COURSE_GRADE_4_F', 'CORE_COURSE_NAME_5_F',
       'CORE_COURSE_GRADE_5_F', 'CORE_COURSE_NAME_6_F',
       'CORE_COURSE_GRADE_6_F', 'SECOND_TERM', 'CORE_COURSE_NAME_1_S',
       'CORE_COURSE_GRADE_1_S', 'CORE_COURSE_NAME_2_S',
       'CORE_COURSE_GRADE_2_S', 'CORE_COURSE_NAME_3_S',
       'CORE_COURSE_GRADE_3_S', 'CORE_COURSE_NAME_4_S',
       'CORE_COURSE_GRADE_4_S', 'CORE_COURSE_NAME_5_S',
       'CORE_COURSE_GRADE_5_S', 'CORE_COURSE_NAME_6_S',
       'CORE_COURSE_GRADE_6_S', 'HOUSING_STS', 'RETURNED_2ND_YR',
       'DISTANCE_FROM_HOME', 'HIGH_SCHL_GPA', 'HIGH_SCHL_NAME',
       'FATHER_HI_EDU_CD', 'FATHER_HI_EDU_DESC', 'MOTHER_HI_EDU_CD',
       'MOTHER_HI_EDU_DESC', 'DEGREE_GROUP_CD', 'DEGREE_GROUP_DESC',
       'FIRST_TERM_ATTEMPT_HRS', 'FIRST_TERM_EARNED_HRS',
       'SECOND_TERM_ATTEMPT_HRS', 'SECOND_TERM_EARNED_HRS', 'GROSS_FIN_NEED',
       'COST_OF_ATTEND', 'EST_FAM_CONTRIBUTION', 'UNMET_NEED'],
      dtype='object')


## PRUNING THE DATA BY REMOVING THE VARIABLES WHICH HAVE LESS EFFECT ON ATTRITION RATE


dm=df.drop(['STDNT_AGE','INTERNATIONAL_STS','STDNT_MINOR','STDNT_TEST_ENTRANCE1', 'STDNT_TEST_ENTRANCE2','FIRST_TERM', 'CORE_COURSE_NAME_1_F',
       'CORE_COURSE_GRADE_1_F', 'CORE_COURSE_NAME_2_F',
       'CORE_COURSE_GRADE_2_F', 'CORE_COURSE_NAME_3_F',
       'CORE_COURSE_GRADE_3_F', 'CORE_COURSE_NAME_4_F',
       'CORE_COURSE_GRADE_4_F', 'CORE_COURSE_NAME_5_F',
       'CORE_COURSE_GRADE_5_F', 'CORE_COURSE_NAME_6_F',
       'CORE_COURSE_GRADE_6_F', 'SECOND_TERM', 'CORE_COURSE_NAME_1_S',
       'CORE_COURSE_GRADE_1_S', 'CORE_COURSE_NAME_2_S',
       'CORE_COURSE_GRADE_2_S', 'CORE_COURSE_NAME_3_S',
       'CORE_COURSE_GRADE_3_S', 'CORE_COURSE_NAME_4_S',
       'CORE_COURSE_GRADE_4_S', 'CORE_COURSE_NAME_5_S',
       'CORE_COURSE_GRADE_5_S', 'CORE_COURSE_NAME_6_S',
       'CORE_COURSE_GRADE_6_S','HIGH_SCHL_NAME',
       'FATHER_HI_EDU_CD','MOTHER_HI_EDU_CD', 'DEGREE_GROUP_CD', 'DEGREE_GROUP_DESC',
       'FIRST_TERM_ATTEMPT_HRS','FIRST_TERM_EARNED_HRS',
       'SECOND_TERM_ATTEMPT_HRS', 'SECOND_TERM_EARNED_HRS', 'GROSS_FIN_NEED','COST_OF_ATTEND'],axis=1)


import matplotlib.pyplot as plt
from sklearn.tree import DecisionTreeClassifier
from sklearn.tree import plot_tree
from sklearn.model_selection import train_test_split
from sklearn.model_selection import cross_val_score
from sklearn.metrics import confusion_matrix,accuracy_score,classification_report
from sklearn.metrics import plot_confusion_matrix
from sklearn import tree


df.isnull().sum()

STUDENT IDENTIFIER             0
STDNT_AGE                      0
STDNT_GENDER                   0
STDNT_BACKGROUND               0
IN_STATE_FLAG                  0
INTERNATIONAL_STS              0
STDNT_MAJOR                    0
STDNT_MINOR                    0
STDNT_TEST_ENTRANCE1        2294
STDNT_TEST_ENTRANCE2         908
STDNT_TEST_ENTRANCE_COMB     518
FIRST_TERM                     0
CORE_COURSE_NAME_1_F           0
CORE_COURSE_GRADE_1_F          0
CORE_COURSE_NAME_2_F          99
CORE_COURSE_GRADE_2_F         99
CORE_COURSE_NAME_3_F         565
CORE_COURSE_GRADE_3_F        565
CORE_COURSE_NAME_4_F        1597
CORE_COURSE_GRADE_4_F       1597
CORE_COURSE_NAME_5_F        2755
CORE_COURSE_GRADE_5_F       2755
CORE_COURSE_NAME_6_F        3272
CORE_COURSE_GRADE_6_F       3272
SECOND_TERM                    0
CORE_COURSE_NAME_1_S         157
CORE_COURSE_GRADE_1_S        232
CORE_COURSE_NAME_2_S         439
CORE_COURSE_GRADE_2_S        439
CORE_COURSE_NAME_3_S        1038
CORE_COURSE_GRADE_3_S       1038
CORE_COURSE_NAME_4_S        2045
CORE_COURSE_GRADE_4_S       2045
CORE_COURSE_NAME_5_S        2950
CORE_COURSE_GRADE_5_S       2950
CORE_COURSE_NAME_6_S        3319
CORE_COURSE_GRADE_6_S       3319
HOUSING_STS                    0
RETURNED_2ND_YR                0
DISTANCE_FROM_HOME            25
HIGH_SCHL_GPA                 53
HIGH_SCHL_NAME                 1
FATHER_HI_EDU_CD             432
FATHER_HI_EDU_DESC             0
MOTHER_HI_EDU_CD             489
MOTHER_HI_EDU_DESC             0
DEGREE_GROUP_CD                0
DEGREE_GROUP_DESC              0
FIRST_TERM_ATTEMPT_HRS         0
FIRST_TERM_EARNED_HRS          0
SECOND_TERM_ATTEMPT_HRS      206
SECOND_TERM_EARNED_HRS       209
GROSS_FIN_NEED                 0
COST_OF_ATTEND                 0
EST_FAM_CONTRIBUTION           0
UNMET_NEED                     0
dtype: int64


len(dm.columns)

14


dm.isnull().sum()

STUDENT IDENTIFIER            0
STDNT_GENDER                  0
STDNT_BACKGROUND              0
IN_STATE_FLAG                 0
STDNT_MAJOR                   0
STDNT_TEST_ENTRANCE_COMB    518
HOUSING_STS                   0
RETURNED_2ND_YR               0
DISTANCE_FROM_HOME           25
HIGH_SCHL_GPA                53
FATHER_HI_EDU_DESC            0
MOTHER_HI_EDU_DESC            0
EST_FAM_CONTRIBUTION          0
UNMET_NEED                    0
dtype: int64


dm=dm.dropna()


dm=pd.get_dummies(dm)


dm.columns

Index(['STUDENT IDENTIFIER', 'STDNT_TEST_ENTRANCE_COMB', 'RETURNED_2ND_YR',
       'DISTANCE_FROM_HOME', 'HIGH_SCHL_GPA', 'EST_FAM_CONTRIBUTION',
       'UNMET_NEED', 'STDNT_GENDER_F', 'STDNT_GENDER_M',
       'STDNT_BACKGROUND_BGD 1', 'STDNT_BACKGROUND_BGD 2',
       'STDNT_BACKGROUND_BGD 3', 'STDNT_BACKGROUND_BGD 4',
       'STDNT_BACKGROUND_BGD 5', 'STDNT_BACKGROUND_BGD 6',
       'STDNT_BACKGROUND_BGD 7', 'STDNT_BACKGROUND_BGD 8', 'IN_STATE_FLAG_N',
       'IN_STATE_FLAG_Y', 'STDNT_MAJOR_Accounting',
       'STDNT_MAJOR_Applied Computer Science', 'STDNT_MAJOR_Art',
       'STDNT_MAJOR_Art Education', 'STDNT_MAJOR_Biology',
       'STDNT_MAJOR_Biology and Secondary Ed', 'STDNT_MAJOR_Chemistry',
       'STDNT_MAJOR_Chemistry and Secondary Ed', 'STDNT_MAJOR_Communication',
       'STDNT_MAJOR_Computer Science - Games',
       'STDNT_MAJOR_Computer Science - Systems',
       'STDNT_MAJOR_Criminal Justice', 'STDNT_MAJOR_Early Admission - Accel',
       'STDNT_MAJOR_Early Childhood Education',
       'STDNT_MAJOR_Earth and Space Science',
       'STDNT_MAJOR_Engineering Studies',
       'STDNT_MAJOR_English Language/Literature',
       'STDNT_MAJOR_English and Secondary Ed', 'STDNT_MAJOR_Exercise Science',
       'STDNT_MAJOR_Finance', 'STDNT_MAJOR_French',
       'STDNT_MAJOR_French with Teacher Cert', 'STDNT_MAJOR_General Business',
       'STDNT_MAJOR_General Studies/AS', 'STDNT_MAJOR_Geology',
       'STDNT_MAJOR_Health Science',
       'STDNT_MAJOR_Health and Physical Education', 'STDNT_MAJOR_History',
       'STDNT_MAJOR_History and Secondary Ed',
       'STDNT_MAJOR_Information Technology',
       'STDNT_MAJOR_Joint Enrollment - Accel', 'STDNT_MAJOR_Liberal Arts',
       'STDNT_MAJOR_Management', 'STDNT_MAJOR_Management Information Systems',
       'STDNT_MAJOR_Marketing', 'STDNT_MAJOR_Mathematics',
       'STDNT_MAJOR_Mathematics and Secondary Ed',
       'STDNT_MAJOR_Middle Grades Education', 'STDNT_MAJOR_Music',
       'STDNT_MAJOR_Music Education', 'STDNT_MAJOR_Music Performance',
       'STDNT_MAJOR_Nursing', 'STDNT_MAJOR_Political Science',
       'STDNT_MAJOR_Pre-Business', 'STDNT_MAJOR_Pre-Engineering/RETP',
       'STDNT_MAJOR_Pre-Nursing', 'STDNT_MAJOR_Psychology',
       'STDNT_MAJOR_Sociology', 'STDNT_MAJOR_Spanish',
       'STDNT_MAJOR_Spanish with Teacher Cert',
       'STDNT_MAJOR_Spec Ed: Gen. Curr. - Reading', 'STDNT_MAJOR_Theatre Arts',
       'STDNT_MAJOR_Theatre Education', 'STDNT_MAJOR_Undeclared',
       'HOUSING_STS_Off Campus', 'HOUSING_STS_On Campus',
       'FATHER_HI_EDU_DESC_College/Beyond', 'FATHER_HI_EDU_DESC_High School',
       'FATHER_HI_EDU_DESC_Middle School/Junior High',
       'FATHER_HI_EDU_DESC_No Information', 'FATHER_HI_EDU_DESC_Other/Unknown',
       'MOTHER_HI_EDU_DESC_College/Beyond', 'MOTHER_HI_EDU_DESC_High School',
       'MOTHER_HI_EDU_DESC_Middle School/Junior High',
       'MOTHER_HI_EDU_DESC_No Information',
       'MOTHER_HI_EDU_DESC_Other/Unknown'],
      dtype='object')


dm=dm.drop(columns=['STDNT_MAJOR_Accounting', 'STDNT_MAJOR_Applied Computer Science',
       'STDNT_MAJOR_Art', 'STDNT_MAJOR_Art Education',
       'STDNT_MAJOR_Biology and Secondary Ed', 'STDNT_MAJOR_Chemistry',
       'STDNT_MAJOR_Chemistry and Secondary Ed', 'STDNT_MAJOR_Communication',
       'STDNT_MAJOR_Computer Science - Games',
       'STDNT_MAJOR_Computer Science - Systems', 'STDNT_MAJOR_Early Admission - Accel',
       'STDNT_MAJOR_Earth and Space Science',
       'STDNT_MAJOR_Engineering Studies',
       'STDNT_MAJOR_English Language/Literature',
       'STDNT_MAJOR_English and Secondary Ed', 'STDNT_MAJOR_Exercise Science', 'STDNT_MAJOR_French',
       'STDNT_MAJOR_French with Teacher Cert', 'STDNT_MAJOR_General Business',
       'STDNT_MAJOR_General Studies/AS', 'STDNT_MAJOR_Geology',
       'STDNT_MAJOR_Health Science',
       'STDNT_MAJOR_Health and Physical Education', 'STDNT_MAJOR_History',
       'STDNT_MAJOR_History and Secondary Ed',
       'STDNT_MAJOR_Information Technology',
       'STDNT_MAJOR_Joint Enrollment - Accel', 'STDNT_MAJOR_Liberal Arts',
       'STDNT_MAJOR_Management', 'STDNT_MAJOR_Management Information Systems',
       'STDNT_MAJOR_Marketing', 'STDNT_MAJOR_Mathematics',
       'STDNT_MAJOR_Mathematics and Secondary Ed',
       'STDNT_MAJOR_Middle Grades Education', 'STDNT_MAJOR_Music',
       'STDNT_MAJOR_Music Education', 'STDNT_MAJOR_Music Performance', 'STDNT_MAJOR_Political Science',
     'STDNT_MAJOR_Pre-Engineering/RETP',
       'STDNT_MAJOR_Sociology', 'STDNT_MAJOR_Spanish',
       'STDNT_MAJOR_Spanish with Teacher Cert',
       'STDNT_MAJOR_Spec Ed: Gen. Curr. - Reading', 'STDNT_MAJOR_Theatre Arts',
       'STDNT_MAJOR_Theatre Education','STDNT_MAJOR_Undeclared'])


len(dm.columns)

39


dm=dm.drop('STUDENT IDENTIFIER',axis=1)


def clean_dataset(dm):
    assert isinstance(dm, pd.DataFrame)
    dm.dropna(inplace=True)
    indices_to_keep = ~dm.isin([np.nan, np.inf, -np.inf]).any(1)
    return df[indices_to_keep].astype(np.float64)


x=dm.drop(columns='RETURNED_2ND_YR')
x


y=dm['RETURNED_2ND_YR']


x.shape

(2832, 37)


y.shape

(2832,)


X_train,X_test,Y_train,Y_test=train_test_split(x,y,random_state=0)


dm.columns

Index(['STDNT_TEST_ENTRANCE_COMB', 'RETURNED_2ND_YR', 'DISTANCE_FROM_HOME',
       'HIGH_SCHL_GPA', 'EST_FAM_CONTRIBUTION', 'UNMET_NEED', 'STDNT_GENDER_F',
       'STDNT_GENDER_M', 'STDNT_BACKGROUND_BGD 1', 'STDNT_BACKGROUND_BGD 2',
       'STDNT_BACKGROUND_BGD 3', 'STDNT_BACKGROUND_BGD 4',
       'STDNT_BACKGROUND_BGD 5', 'STDNT_BACKGROUND_BGD 6',
       'STDNT_BACKGROUND_BGD 7', 'STDNT_BACKGROUND_BGD 8', 'IN_STATE_FLAG_N',
       'IN_STATE_FLAG_Y', 'STDNT_MAJOR_Biology',
       'STDNT_MAJOR_Criminal Justice', 'STDNT_MAJOR_Early Childhood Education',
       'STDNT_MAJOR_Finance', 'STDNT_MAJOR_Nursing',
       'STDNT_MAJOR_Pre-Business', 'STDNT_MAJOR_Pre-Nursing',
       'STDNT_MAJOR_Psychology', 'HOUSING_STS_Off Campus',
       'HOUSING_STS_On Campus', 'FATHER_HI_EDU_DESC_College/Beyond',
       'FATHER_HI_EDU_DESC_High School',
       'FATHER_HI_EDU_DESC_Middle School/Junior High',
       'FATHER_HI_EDU_DESC_No Information', 'FATHER_HI_EDU_DESC_Other/Unknown',
       'MOTHER_HI_EDU_DESC_College/Beyond', 'MOTHER_HI_EDU_DESC_High School',
       'MOTHER_HI_EDU_DESC_Middle School/Junior High',
       'MOTHER_HI_EDU_DESC_No Information',
       'MOTHER_HI_EDU_DESC_Other/Unknown'],
      dtype='object')


from sklearn.ensemble import RandomForestClassifier
classifier=RandomForestClassifier(max_depth=7)
classifier.fit(x,y)

RandomForestClassifier(max_depth=7)


y_pred=classifier.predict(X_test)
print(confusion_matrix(Y_test,y_pred))
print(accuracy_score(Y_test,y_pred))
print(classification_report(Y_test,y_pred))

[[  3 148]
 [  0 557]]
0.7909604519774012
              precision    recall  f1-score   support

           0       1.00      0.02      0.04       151
           1       0.79      1.00      0.88       557

    accuracy                           0.79       708
   macro avg       0.90      0.51      0.46       708
weighted avg       0.83      0.79      0.70       708


# As you can see the tree is highly biased to the majority value as the data is class imbalanced
#so i am using UNDER SAMPLING inorder to tackle the situation


y.value_counts()

1    2243
0     589
Name: RETURNED_2ND_YR, dtype: int64


count_class_1,count_class_0=dm.RETURNED_2ND_YR.value_counts()
count_class_0,count_class_1

(589, 2243)


dm_class_0=dm[dm['RETURNED_2ND_YR']==0]
dm_class_1=dm[dm['RETURNED_2ND_YR']==1]


dm_class_0.shape

(589, 38)


dm_class_1.shape

(2243, 38)


dm_class_1_under=dm_class_1.sample(count_class_0)
dm_test_under=pd.concat([dm_class_1_under,dm_class_0],axis=0)


dm_test_under.shape

(1178, 38)


x=dm_test_under.drop('RETURNED_2ND_YR',axis=1)


y=dm_test_under['RETURNED_2ND_YR']


X_train,X_test,Y_train,Y_test=train_test_split(x,y,random_state=0,stratify=y)


from sklearn.ensemble import RandomForestClassifier
classifier=RandomForestClassifier(n_estimators=100,max_depth=4)
classifier.fit(x,y)

RandomForestClassifier(max_depth=4)


y_pred=classifier.predict(X_test)
print(confusion_matrix(Y_test,y_pred))
print(accuracy_score(Y_test,y_pred))
print(classification_report(Y_test,y_pred))

[[86 62]
 [59 88]]
0.5898305084745763
              precision    recall  f1-score   support

           0       0.59      0.58      0.59       148
           1       0.59      0.60      0.59       147

    accuracy                           0.59       295
   macro avg       0.59      0.59      0.59       295
weighted avg       0.59      0.59      0.59       295


classifier=RandomForestClassifier(n_estimators=100,max_depth=7)
classifier.fit(x,y)

RandomForestClassifier(max_depth=7)


y_pred=classifier.predict(X_test)
print(confusion_matrix(Y_test,y_pred))
print(accuracy_score(Y_test,y_pred))
print(classification_report(Y_test,y_pred))

[[107  41]
 [ 22 125]]
0.7864406779661017
              precision    recall  f1-score   support

           0       0.83      0.72      0.77       148
           1       0.75      0.85      0.80       147

    accuracy                           0.79       295
   macro avg       0.79      0.79      0.79       295
weighted avg       0.79      0.79      0.79       295


#THIS MODEL HAS A F1 SCORE OF 79% WHICH MAKES THIS AS THE BEST ONE


clf=DecisionTreeClassifier(random_state=0,max_depth=7)
clf=clf.fit(X_train,Y_train)


plot_confusion_matrix(clf,X_test,Y_test,display_labels=['HAVE NOT RETURNED','RETURNED'])

<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x14538368dc0>


clf.score(X_test,Y_test)

0.4576271186440678


# SO RANDOM FOREST IS THE MOST SUITABLE ONE


from sklearn.tree import export_graphviz
# Export as dot file
export_graphviz(classifier.estimators_[1], out_file='tree.dot', 
                feature_names = x.columns,
                class_names =['HAVE NOT RETURNED','RETURNED'] ,
                rounded = True, proportion = False, 
                 filled = True)


from subprocess import call
call(['dot', '-Tpng', 'tree.dot', '-o', 'tree.png', '-Gdpi=600'])

# Display in jupyter notebook
from IPython.display import Image
Image(filename = 'tree.png')

	Count	Unique	Miss_per	Min	Max	Mean	Median	Std	Var	Q1	Q3	q99
STUDENT IDENTIFIER	3400.0	3400.0	0.000000	7755837.0	8037098.0	7.897155e+06	7899052.5	80834.212657	6.534170e+09	7827039.25	7964008.00	8034112.56
STDNT_AGE	3400.0	11.0	0.000000	16.0	26.0	1.799412e+01	18.0	0.552052	3.047609e-01	18.00	18.00	19.00
STDNT_TEST_ENTRANCE1	1106.0	24.0	0.674706	8.0	32.0	2.066004e+01	20.0	3.234734	1.046351e+01	18.00	22.00	30.00
STDNT_TEST_ENTRANCE2	2492.0	110.0	0.267059	350.0	1490.0	9.650923e+02	990.0	209.088633	4.371806e+04	900.00	1100.00	1350.90
STDNT_TEST_ENTRANCE_COMB	2882.0	25.0	0.152353	530.0	1510.0	9.977585e+02	990.0	156.268221	2.441976e+04	910.00	1110.00	1340.00
FIRST_TERM	3400.0	6.0	0.000000	200508.0	201008.0	2.007770e+05	200808.0	170.001298	2.890044e+04	200608.00	200908.00	201008.00
SECOND_TERM	3400.0	6.0	0.000000	200602.0	201102.0	2.008710e+05	200902.0	170.001298	2.890044e+04	200702.00	201002.00	201102.00
RETURNED_2ND_YR	3400.0	2.0	0.000000	0.0	1.0	7.873529e-01	1.0	0.409240	1.674775e-01	1.00	1.00	1.00
DISTANCE_FROM_HOME	3375.0	72.0	0.007353	0.0	5932.0	9.982519e+01	69.0	235.196984	5.531762e+04	69.00	138.00	666.00
HIGH_SCHL_GPA	3347.0	201.0	0.015588	0.0	4.0	3.202738e+00	3.2	0.455337	2.073318e-01	2.87	3.56	4.00
FATHER_HI_EDU_CD	2968.0	5.0	0.127059	1.0	4.0	2.560647e+00	3.0	0.681273	4.641333e-01	2.00	3.00	4.00
MOTHER_HI_EDU_CD	2911.0	5.0	0.143824	1.0	4.0	2.571281e+00	3.0	0.611729	3.742128e-01	2.00	3.00	4.00
FIRST_TERM_ATTEMPT_HRS	3400.0	12.0	0.000000	9.0	21.0	1.398588e+01	14.0	1.554286	2.415805e+00	13.00	15.00	18.00
FIRST_TERM_EARNED_HRS	3400.0	22.0	0.000000	0.0	21.0	1.220500e+01	13.0	2.986294	8.917951e+00	11.00	15.00	17.00
SECOND_TERM_ATTEMPT_HRS	3194.0	22.0	0.060588	2.0	23.0	1.428585e+01	14.0	2.104387	4.428444e+00	13.00	16.00	19.00
SECOND_TERM_EARNED_HRS	3191.0	24.0	0.061471	0.0	23.0	1.251113e+01	13.0	3.470492	1.204431e+01	11.00	15.00	19.00
GROSS_FIN_NEED	3400.0	927.0	0.000000	0.0	2124900.0	3.034025e+05	0.0	483625.450816	2.338936e+11	0.00	599580.00	1418100.00
COST_OF_ATTEND	3400.0	120.0	0.000000	0.0	2124900.0	5.515335e+05	150540.0	606440.325202	3.677699e+11	0.00	1192815.00	1489080.00
EST_FAM_CONTRIBUTION	3400.0	1236.0	0.000000	0.0	5999940.0	3.446899e+05	0.0	787187.774837	6.196646e+11	0.00	324180.00	3905646.60
UNMET_NEED	3400.0	1197.0	0.000000	-1212072.0	1632660.0	7.222052e+04	0.0	279567.921396	7.815822e+10	0.00	187615.50	883023.60

	STUDENT IDENTIFIER	STDNT_AGE	STDNT_GENDER	STDNT_BACKGROUND	IN_STATE_FLAG	INTERNATIONAL_STS	STDNT_MAJOR	STDNT_MINOR	STDNT_TEST_ENTRANCE1	STDNT_TEST_ENTRANCE2	...	DEGREE_GROUP_CD	DEGREE_GROUP_DESC	FIRST_TERM_ATTEMPT_HRS	FIRST_TERM_EARNED_HRS	SECOND_TERM_ATTEMPT_HRS	SECOND_TERM_EARNED_HRS	GROSS_FIN_NEED	COST_OF_ATTEND	EST_FAM_CONTRIBUTION	UNMET_NEED
0	7808615	18	F	BGD 1	Y	N	Undeclared	N	NaN	1150.0	...	B	Bachelors	16	16	14.0	14.0	0	0	0	0.0
11	7884573	18	F	BGD 1	Y	N	Undeclared	N	NaN	1050.0	...	B	Bachelors	16	16	16.0	16.0	0	1189920	5163240	-326880.0
17	8004193	17	M	BGD 1	N	N	Biology	N	NaN	1040.0	...	B	Bachelors	14	14	14.0	14.0	1083000	1324500	241500	476580.0
23	8007842	18	F	BGD 1	Y	N	Pre-Nursing	N	NaN	580.0	...	B	Bachelors	15	15	15.0	15.0	0	0	0	0.0
36	7799972	18	F	BGD 1	Y	N	Undeclared	N	NaN	620.0	...	B	Bachelors	18	18	14.0	14.0	0	0	0	0.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
3394	7960053	18	F	BGD 1	Y	N	Theatre Arts	N	NaN	1010.0	...	B	Bachelors	13	3	NaN	NaN	0	0	0	0.0
3395	7948112	18	F	BGD 4	N	N	Music Performance	N	NaN	800.0	...	B	Bachelors	18	18	16.0	13.0	0	0	0	0.0
3396	8023055	18	F	BGD 1	Y	N	Biology	N	20.0	NaN	...	B	Bachelors	13	9	NaN	NaN	0	0	0	0.0
3397	7926915	18	F	BGD 4	Y	N	Biology	N	NaN	1020.0	...	B	Bachelors	14	6	NaN	NaN	0	945840	1288980	0.0
3399	7928405	18	M	BGD 2	Y	N	Art	Creative Writing	NaN	890.0	...	B	Bachelors	12	9	13.0	6.0	0	0	0	0.0

	Count	Unique	Miss_per	Freq_Level	freq_cat_Per
STDNT_GENDER	3400	2	0	F	0.629412
STDNT_BACKGROUND	3400	8	0	BGD 1	0.612353
IN_STATE_FLAG	3400	2	0	Y	0.900882
INTERNATIONAL_STS	3400	2	0	N	0.992059
STDNT_MAJOR	3400	54	0	Undeclared	0.135294
STDNT_MINOR	3400	37	0	N	0.929412
CORE_COURSE_NAME_1_F	3400	69	0	ENGL 1101	0.183529
CORE_COURSE_GRADE_1_F	3400	7	0	C	0.301176
CORE_COURSE_NAME_2_F	3301	75	0.0291176	ENGL 1101	0.216471
CORE_COURSE_GRADE_2_F	3301	8	0.0291176	B	0.393824
CORE_COURSE_NAME_3_F	2835	75	0.166176	ENGL 1101	0.167059
CORE_COURSE_GRADE_3_F	2835	7	0.166176	B	0.339706
CORE_COURSE_NAME_4_F	1803	69	0.469706	ENGL 1101	0.0914706
CORE_COURSE_GRADE_4_F	1803	7	0.469706	A	0.321765
CORE_COURSE_NAME_5_F	645	54	0.810294	ENGL 1101	0.0238235
CORE_COURSE_GRADE_5_F	645	5	0.810294	A	0.147647
CORE_COURSE_NAME_6_F	128	25	0.962353	ITDS 2749	0.00588235
CORE_COURSE_GRADE_6_F	128	5	0.962353	A	0.0323529
CORE_COURSE_NAME_1_S	3243	67	0.0461765	ENGL 1102	0.184412
CORE_COURSE_GRADE_1_S	3168	9	0.0682353	C	0.266471
CORE_COURSE_NAME_2_S	2961	70	0.129118	ENGL 1102	0.181471
CORE_COURSE_GRADE_2_S	2961	7	0.129118	B	0.328824
CORE_COURSE_NAME_3_S	2362	69	0.305294	ENGL 1102	0.119118
CORE_COURSE_GRADE_3_S	2362	7	0.305294	A	0.292647
CORE_COURSE_NAME_4_S	1355	65	0.601471	ENGL 1102	0.0582353
CORE_COURSE_GRADE_4_S	1355	6	0.601471	A	0.226471
CORE_COURSE_NAME_5_S	450	52	0.867647	MUSC 1100	0.0129412
CORE_COURSE_GRADE_5_S	450	6	0.867647	A	0.0891176
CORE_COURSE_NAME_6_S	81	27	0.976176	POLS 1101	0.00294118
CORE_COURSE_GRADE_6_S	81	4	0.976176	A	0.0191176
HOUSING_STS	3400	2	0	Off Campus	0.582647
HIGH_SCHL_NAME	3399	440	0.000294118	SCHOOL 11	0.0732353
FATHER_HI_EDU_DESC	3400	5	0	High School	0.392941
MOTHER_HI_EDU_DESC	3400	5	0	College/Beyond	0.441176
DEGREE_GROUP_CD	3400	3	0	B	0.995294
DEGREE_GROUP_DESC	3400	3	0	Bachelors	0.995294

	STDNT_TEST_ENTRANCE_COMB	DISTANCE_FROM_HOME	HIGH_SCHL_GPA	EST_FAM_CONTRIBUTION	UNMET_NEED	STDNT_GENDER_F	STDNT_GENDER_M	STDNT_BACKGROUND_BGD 1	STDNT_BACKGROUND_BGD 2	STDNT_BACKGROUND_BGD 3	...	FATHER_HI_EDU_DESC_College/Beyond	FATHER_HI_EDU_DESC_High School	FATHER_HI_EDU_DESC_Middle School/Junior High	FATHER_HI_EDU_DESC_No Information	FATHER_HI_EDU_DESC_Other/Unknown	MOTHER_HI_EDU_DESC_College/Beyond	MOTHER_HI_EDU_DESC_High School	MOTHER_HI_EDU_DESC_Middle School/Junior High	MOTHER_HI_EDU_DESC_No Information	MOTHER_HI_EDU_DESC_Other/Unknown
0	1150.0	150.0	4.00	0	0.0	1	0	1	0	0	...	0	0	0	1	0	0	0	0	1	0
1	1190.0	69.0	2.89	785760	459300.0	1	0	1	0	0	...	1	0	0	0	0	1	0	0	0	0
2	1030.0	150.0	3.39	0	0.0	0	1	1	0	0	...	0	1	0	0	0	0	0	0	0	1
3	1220.0	150.0	2.93	0	0.0	0	1	1	0	0	...	1	0	0	0	0	1	0	0	0	0
4	1190.0	69.0	3.86	519840	278340.0	1	0	1	0	0	...	1	0	0	0	0	0	1	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
3395	790.0	871.0	3.93	0	0.0	1	0	0	0	0	...	0	0	0	1	0	0	0	0	1	0
3396	950.0	69.0	2.69	0	0.0	1	0	1	0	0	...	0	0	0	1	0	0	0	0	1	0
3397	1030.0	90.0	2.76	1288980	0.0	1	0	0	0	0	...	0	0	0	0	1	0	1	0	0	0
3398	990.0	0.0	3.56	0	-231720.0	1	0	0	0	1	...	0	1	0	0	0	1	0	0	0	0
3399	870.0	69.0	2.63	0	0.0	0	1	0	1	0	...	1	0	0	0	0	1	0	0	0	0

	RETURNED_2ND_YR	EST_FAM_CONTRIBUTION
0	0	315265.643154
1	1	352636.787449

	RETURNED_2ND_YR	UNMET_NEED
0	0	90679.460581
1	1	67235.153455

	RETURNED_2ND_YR	HIGH_SCHL_GPA
0	0	3.131802
1	1	3.221768

	RETURNED_2ND_YR	FIRST_TERM_EARNED_HRS
0	0	11.809129
1	1	12.311916